Havuz

Zero-overhead, terminal-native local-LLM orchestrator. Run any engine, any model via the same proxy.

An open-source Digital Worker platform for reliable execution, continuous co-evolution, and building Enterprise AI assets.

The operational superset of Pi Coding Agent — everything Pi, plus observability, governance, recovery, evaluation and multi-agent orchestrat

Local-first persistent memory for AI coding agents (Claude Code, Cursor, Codex) over MCP. Decisions, lessons and facts live in one SQLite fi

llama.cpp/ik_llama.cpp launcher: loads big MoE models across mismatched multi-GPU rigs by exact VRAM math.

Automated alignment adjustment for LLMs — direct steering, LoRA, and MoE expert-granular abliteration, optimized via multi-objective Optuna

Monet — Multi-engine mission control for coding agents (Claude Code and Codex today). Browse, search, and drive your agent sessions from a n

Full-stack open-source AI engine for building language models — tokenizer training, transformer architecture, cognitive reasoning and chat p

🦞 ResearchClawBench: Evaluating AI Agents for Automated Research from Re-Discovery to New-Discovery

面向 AI 编程智能体的任务包工作流工具，用可验证的范围、证据和 Git 门禁管理长期改动。(Task-package workflow for AI coding agents with verifiable scope, evidence, and Git gates.)

Your AI intranet: network the computers you already own for inference and training.

Refusal removal as a loadable KV-cache graft — zero weight modification, no refusal-direction projection, fully reversible.

基于 DeepSeek Harness（DSH）的稳定桌面端，集成Git、内置浏览器与记忆功能 | DeepSeek Harness desktop GUI with local workspaces, Git, browser and memory.

YouTube transcript skills for Claude Code, Cursor and Codex. Agent Skills to get transcripts, search YouTube, read channels and playlists, a

High-performance Go proxy gateway for 9Router LLM routing — 32K RPS, 42MB RAM, CGO-free

鲸鲸 · 让 DeepSeek 用你的微信账号陪聊（Windows 桌面端 + WPF 启动器）。WeChat chat bot powered by DeepSeek/Qwen.

LLM experiment reproducibility toolkit

A 14M-parameter LLM keeps a virtual fish tank alive on an $8 ESP32-S3 — distilled from a 26B teacher into 7.56 MB, no network.

An edge-native Multimodal Android Agent that integrates multimodal perception, memory and action

[Public preview] Externally scored agentic ML research benchmark: 60 tasks, real competition ground truth. Open protocol, operated evaluatio

Local proxy that compresses your LLM API requests so you pay less, with no change to the answers. Trims wasted tokens from prompts, history,

Official code for the paper "Memora: A Harmonic Memory Representation Balancing Abstraction and Specificity"

Open-source app builder engine — intent to working app

🧬 Symbio（共生）- AI Infra 级多 Agent 协同框架 | 33 Killer Features: 动态 DAG, 本体化记忆, 防过早完成, 数据飞轮, 神经符号安全, OpenTelemetry 可观测, HITL 人类介入, MCP 原生支持, 语义缓存,

One API for every LLM: 170+ providers through one client, with a Rust core and native bindings for every major language.

TOMC Memory: CPU context compression and task memory for Codex, Claude Code, Cursor, TRAE, WorkBuddy and other MCP clients.

An agentic system that auto-optimizes LLM workloads on AMD GPUs.

A multi-model Muse: personal AI voice assistant on ESP32 - wake word or knob, any OpenAI-compatible brain (STT/chat/TTS mix-and-match), stre

Custom APRL machine-learning algorithm + a curated ML knowledge base of 923 papers, lectures, and explainers. NumPy classifier/regressor, te

A living AI-agent office on your desktop wallpaper — Claude Code agents that walk, work, delegate, learn & hold meetings. Per-agent swappabl

Tactical AI Workspace Monitor & EDR

Established strong JEV based natural language to SQL performance and well-covered semantic operators for PostgreSQL with optimized cost and

declarative self-improving language-model programs for Elixir 😇

Open-source memory runtime for AI agents — reproducible, provenance-tagged context bundles instead of query-time retrieval. Apache-2.0, self

106 social media skills for AI agents - strategy, writing, video, design, platform growth, publishing, and analytics. Works with Claude, Cur

Spec-driven autonomous engineering workflow for AI coding agents: recon → plan → build → validate → ship — with TDD, adversarial validation,

Jev-Mem: System-One Controlled Agentic Memory

Turn your documents into a Markdown wiki you can read, that answers with citations. Works with any LLM provider, Ollama included. CROW class

繁體中文（台灣）個人資料去識別化工具，讓業務文件可以安全地送進 AI 處理 / Reversible Traditional-Chinese (Taiwan) PII de-identification for LLM workflows — fully local

The lossless, headless coding agent harness. One 7 MB static binary, four tools, plain stdin/stdout, no silent cuts.

Automated quality checks that nudge AI coding agents toward better habits

Everything that *went* wrong with Claude.

The open-source AI agent SDK for WhatsApp. Bring your own model, your own number, your own data. Local-first, no forced cloud, no vendor loc

llama.cpp fork for significantly improved performance on Ampere generation Nvidia cards (especially RTX 3090 / 3090 Ti): example: 110+ tok/s

Codex 风格的推理等级滑条（DeepSeek Harness 插件）：可拖动连续滑条 + 高档位渐变星尘特效

Interactive desktop app with 58 labs covering ML fundamentals, transformers, RAG, agents, fine-tuning and quantization and more.

Private AI on premise

Real-time multimodal AI pipelines on Apple Silicon

Self-hosted web scraper, crawler and document parser for AI agents: URLs, PDFs with OCR and Office files to clean Markdown/JSON. Firecrawl-

Your AI's guard dog. Doberman sits at runtime, gating every input, output and tool call to stop unsafe or unintended actions before they exe